Adoption of autonomous driving technology: An integration of innovation diffusion theory and motivation theory

Abstract

Background

Autonomous vehicles (AVs) are poised to become transformative technologies in modern or smart cities. However, consumers’ reluctance to adopt AVs remains a significant barrier.

Objective

This study aims to investigate three questions. First, which innovation characteristics of AVs most influence consumer adoption? Secondly, how do perceived benefits and risks shape adoption decisions, and are risks more influential than benefits? Thirdly, to test the mediating effects of consumer value perceptions, as well as other fatal contingent mechanisms in the above causal relationships.

Methods

A structural survey based on questionnaires was conducted, and a valid sample containing 504 respondents was finally collected. To explore the causal relationships, the regression method was employed to analyze the relationships among key factors affecting AVs diffusion.

Results

Three key findings were indicated. First, ease of use, compatibility, and relative advantage significantly influence AV adoption. Second, perceived usefulness, hedonism, and risk all affect adoption, with perceived risk showing the strongest negative impact. Third, these value perceptions mediate the relationship between innovation characteristics and adoption intention, outlining the cognitive pathway of AV diffusion. Additionally, subjective norms moderate the impact of perceived risk, suggesting a crowd effect in adoption decisions for emerging technologies.

Conclusions

These findings contribute to the advancement of human–machine interaction theory and provide valuable insights for developing policies and strategies to promote AV adoption.

Keywords

intention of AV adoption innovation diffusion theory motivation theory perceived risk

Introduction

Autonomous vehicles (AVs), as a disruptive innovation, have recently attracted widespread attention and discourse within both academic and industry circles. According to a report by IDATE, AVs represent a potential market of 55 million units over the next two decades, with Asian countries projected to lead global AV sales by 2040.¹ The core technologies enabling AVs—such as lidar, sensor fusion, artificial intelligence, and 5G connectivity—are becoming increasingly accessible, while regulatory frameworks are gradually evolving to support their deployment.^2,3 Already, consumer‐grade semiautonomous vehicles have been marketed. Thus, automated driving technology is progressively entering public awareness, requiring users to adapt to a new mode of human–machine shared driving. Throughout the evolution of this technology, nowadays, Level 3 of AVs is fast launching in the market. In China, the penetration rate of Level 3 autonomous driving has reached 10%. With rapid advances in technology innovation and supporting infrastructure, AVs are projected to occupy 25% of the global private vehicle market by 2040.⁴

The promotion and utilization of autonomous driving technology are poised to yield substantial social and economic benefits, including the alleviation of urban traffic congestion, the substantial enhancement of road safety, the increase in overall transport efficiency, and the reduction in social logistics costs.^5–7 Moreover, at the individual level, it offers technological solutions that have the potential to enhance the quality of life. For example, autonomous driving can provide a more convenient way of traveling for elderly people and people with mobility impairments, enhancing their travel autonomy and quality of life.⁸

However, public acceptance and willingness to adopt the technology remain significant concerns, which continue to limit its application and diffusion.^9–11 The most important concern should be that users have worries about the safety of autonomous driving technology and the infringement of their privacy data. For the currently available Level 3 driving system named with conditional automation, there are radical differences from Level 2, featured by assistant driving, and Level 4, featured by high automation. Level 3 represents a critical transition phase from “human–machine shared driving” to “machine-dominant” operation. Its core technical features revolve around system reliability in perception, decision-making, and execution, as well as safety in human–machine interaction. Consumers’ risk perceptions towards Level 3 autonomous driving technology include several aspects, such as safety, abuse of personal privacy information, legal liability, technical reliability, and cost-effectiveness. The relative importance of the benefits and risks is still unknown in the process of AVs innovation diffusion.

Theoretical foundation and literature review

Innovation diffusion theory

Innovation diffusion theory (IDT), proposed by Rogers in 1962, is a well-established framework for examining the process of consumer acceptance of new technologies.¹² IDT suggests that the process of diffusion of innovations consists of five phases: knowledge, persuasion, decision, implementation, and confirmation.¹² Rogers further emphasized that perceived innovation characteristics—relative advantage, compatibility, complexity, trialability, and observability—are critical determinants of adoption decisions. Among these, relative advantage, compatibility, and complexity are the most influential factors driving consumer acceptance of new technologies and are consistently associated with innovation adoption.¹³ Moore and Benbasat subsequently suggested renaming “complexity” as “ease of use,” a modification widely adopted.^14,15

The theory has been widely applied in the fields of technology adoption, marketing, and social change.¹⁶ Current research has identified two core streams: the first takes a comprehensive approach to studying the continuous diffusion process of innovations, examining how innovations evolve and propagate through various stages within social systems¹⁷; the second examines the impact of perceived innovation characteristics, analyzing how these characteristics influence users’ decisions in different contexts.¹⁸ As an emerging AI technology, AVs are particularly suited for the application of this theory. However, since AVs have not yet been fully commercialized, their diffusion process, according to IDT’s stages, remains at the “persuasion stage”—where potential users begin to learn about the technology and form subjective perceptions, but have not yet moved to large-scale adoption. Against this backdrop, most research applying IDT to AVs focuses on “perceived innovation characteristics.” For instance, Yuen et al. argued that IDT is effective in identifying the diffusion characteristics of AVs and found that relative advantage and compatibility positively influence perceived usefulness.⁴ Similarly, Farzina et al. integrated perceived innovation characteristics with the Unified Theory of Acceptance and Use of Technology (UTAUT) to identify potential factors influencing the acceptance of fully automated AVs.¹⁹

Motivation theory

Motivation theory has been widely used to explain the behavior of individuals adopting information technology.²⁰ Ryan and Deci classified the motivation behind an individual’s behavior into extrinsic and intrinsic motivation.²¹ In this context, extrinsic motivation refers to individuals acting because they perceive that behavior helps achieve an external goal (e.g., performance enhancement), whereas intrinsic motivation refers to individuals acting owing to an inherent interest in the behavior itself, independent of external rewards.

Many scholars have explored the adoption of autonomous driving technology from the perspective of motivational theory, emphasizing the dual roles of extrinsic and intrinsic motivation. Perceived usefulness, a core construct in the Technology Acceptance Model (TAM), has been widely validated as a typical indicator of extrinsic motivation,²² as it reflects users’ cognitive evaluations of the practical benefits and performance enhancements that the technology may offer. In contrast, perceived hedonism, derived from UTAUT2,²³ constitutes a central dimension of intrinsic motivation, capturing users’ direct experiences of pleasure, novelty, and emotional satisfaction during their interactions with the technology. For example, Seuwou developed a model of AV technology acceptance from a motivational perspective, combining the UTAUT2 model with other determinants (AVTAM).²⁴ Hegner et al. considered both internal and external motivations when investigating the acceptance of AVs, where perceived usefulness was regarded as intrinsic motivation and perceived hedonism as extrinsic motivation.²⁵

The opposite side of motivation should not be neglected when exploring the factors of adoption decision, such as perceived risks. Perceived risk was first systematically proposed by Bauer (1960), refers to the psychological reactions of individuals on the basis of their perceptions of uncertainty and potential negative consequences associated with new products or technologies. Risk perception encompasses the cognitive process through which individuals assess and avoid harmful environmental conditions, emphasizing the role of intuitive judgments in evaluating potential threats.²⁶ As such, user-perceived risk is highly subjective. In the context of the growing popularity of AI and AVs, perceived risk has emerged as a significant barrier to individual decision-making in the technology adoption process.²⁷ As innovative products with considerable technological complexity, AVs have potential impacts on users’ lifestyles, social interactions, and even personal safety, making risk a critical consideration in consumer acceptance of the technology. For example, many consumers are concerned that errors in autonomous driving technology could lead to traffic accidents, thereby increasing safety and performance risks.^28,29 On the other hand, autonomous driving technologies typically rely on extensive data collection and processing, including information on driver behavior, vehicle paths, and environmental factors.³⁰ This reliance on data may exacerbate consumers’ concerns about privacy.

Although this topic has attracted considerable attention, there are still some limitations in the research paradigm of the literature. First, while perceived value is widely recognized as a critical determinant of user adoption of technology,³¹ existing literature offers insufficient insight into how consumers interpret the innovation value of autonomous driving technology. Therefore, in the context of AV technology diffusion, it is essential to investigate the value perceptions elicited by specific innovation characteristics and to empirically assess their interrelationships. Such an inquiry will help identify the key innovation attributes that effectively facilitate the successful market penetration of AV technology.

Second, the criteria by which consumers evaluate the value of autonomous driving technology need to be systematically constructed. In contrast to traditional technologies—where decision-making authority remains entirely with humans—autonomous driving systems delegate part, or even all, of that authority to machines, thereby introducing novel and potentially significant risks.¹¹ Accordingly, there is a need to construct a value perception framework grounded in two dimensions: perceived benefits and perceived risks. Clarifying the relative weight of these dimensions in consumer decision-making processes will provide critical insights into the cognitive logic that underlies AV adoption behavior.

Third, the decision-making mechanisms that drive consumer adoption of AV technology remain insufficiently understood. Specifically, what are the direct effects of technological innovation characteristics and perceived value on adoption intention? Furthermore, does consumer value perception function as a significant mediating variable in this process? To address these questions, it is necessary to empirically examine both the direct and indirect pathways of influence, thereby uncovering the underlying psychological and cognitive mechanisms guiding AV adoption decisions.

In response to the limitations of existing research, this study proposes a research framework based on innovation diffusion theory, motivation theory, and perceived risk theory. It examines the direct effects of technological innovation characteristics and user value perception on adoption intention, and further tests the mediating role of value perception. Additionally, previous studies have primarily treated subjective norms as independent variables, focusing on their direct impact on adoption intentions,^32,33 while paying limited attention to their diverse moderating roles in consumers’ complex psychological decision-making processes. Considering the influence of social norms in the Chinese context, the framework incorporates social influence factors to provide a more comprehensive understanding of user adoption mechanisms for autonomous driving technology. The proposed mechanisms are tested via empirical data.

Research model and hypotheses

Model construction

Autonomous vehicles represent an innovation that has fundamentally transformed the way passengers and goods are transported. Therefore, the IDT is a suitable theoretical framework for this study. IDT posits that the adoption of innovations is driven by their perceived attributes, underscoring the influence of technological features on user decision-making. In contrast, Motivation theory adopts a consumer-centric lens, explaining how users’ positive perceptions correlate with their acceptance of new technologies. Despite the advantages of autonomous driving technology, its technical complexity entails substantial potential risks. Within this context, we employ both the positive motivational factors (reflecting the perceived benefits) and the opposite motivational factors, which are perceived risks, so as to form a “benefit-risk” dual perception framework that shapes user decisions.^34,35 Accordingly, this study proposes an integrated framework based on IDT and motivation theory to explore the decisive mechanism of adoption intention of AVs. The model focuses on both the direct effects of technology attributes and consumer motivations on adoption intention and the potential mediating effects played by motivations. Therefore, we can illustrate the diffusion process of AV technology. The research model is given below (see Figure 1).

Figure 1.

The theoretical model.

Effects of perceived value on adoption intention

Consistent with Davis et al. (1989), perceived usefulness (PU) in this study refers to the extent to which individuals believe that using a particular system will enhance their job performance.²² As individuals’ expectations of a system increase, their perception of its value tends to improve. Previous research consistently shows that user acceptance of innovative technologies—across domains such as consumer technology, information systems, and others—correlates with perceptions of the technology’s usefulness.^36,37 In the context of AVs, perceived usefulness similarly predicts users’ willingness to adopt the technology. For example, J. Park et al. found that perceived usefulness significantly influences the willingness to use AVs, with demographic variables such as age, marital status, family size, and education level moderating this relationship.³⁸ Xiao and Goulias further explored both demographic and contextual factors, including travel behavior, and identified perceived usefulness as a key determinant of adoption intention.³⁹ Based on these findings, the following hypotheses are proposed:

Perceived usefulness has a positive effect on the intention of AV adoption.

Perceived hedonism refers to the pleasure or enjoyment an individual derives from using a particular technology.²³ In the context of AVs, perceived hedonism enhances user acceptance by offering novel and complex sensory experiences. Individuals with hedonistic tendencies are more likely to seek innovative and unique experiences, increasing their willingness to adopt emerging technologies such as AVs.⁴⁰ For instance, Huang found that hedonic factors significantly influence the intention to use AVs, particularly regarding psychological factors shaping users’ willingness to adopt the technology.⁴¹ Moreover, perceived hedonic enjoyment strengthens users’ emotional connection and satisfaction, which, in turn, boosts their intention to use AVs. Positive experiences during AV use increase the likelihood of continued usage.⁴² Based on these findings, we propose the following hypothesis:

Perceived hedonism has a positive effect on the intention of AV adoption.

In consumer behavior research, perceived risk refers to the expectation of potential loss under uncertainty.⁴³ Numerous studies have indicated that while respondents acknowledge the benefits of AVs, they also express significant concerns about the associated risks.^44,45 Perceived risk can erode users’ trust in AVs, thereby diminishing their willingness to adopt the technology. Autonomous driving technology presents complex technical, legislative, and ethical challenges,² including risks such as traffic accidents due to technological failures and privacy breaches, both of which may lead users to question the technology’s safety and reliability.⁴⁶ These concerns, in turn, reduce acceptance of the technology. Moreover, perceived risk amplifies anxiety during the decision-making process,⁴⁴ as users worry about the system’s performance in complex traffic situations and its ability to handle unforeseen circumstances. Thus, the perceived risks in this study include privacy, safety, and performance risks. Based on these considerations, the following hypotheses are proposed:

Perceived risk has a negative effect on the intention of AV adoption.

Effects of innovation characteristics on adoption intention

Ease of use refers to the extent to which users find a system easy to operate.⁴⁷ In the Technology Acceptance Model (TAM), many studies have found evidence on the positive relationship between ease of use and intention to use²² of a technology. For instance, Ma et al. confirmed that perceived ease of use positively influences users’ behavioral intention to adopt ChatGPT.⁴⁸ Pillai et al. investigated the factors affecting students’ adoption of AI robots in educational environments and revealed that perceived ease of use had a significant positive effect on their behavioral intention to use the technology.⁴⁹

Existing studies have revealed the impact of ease of use on the intention to use from various perspectives. For an autonomous driving system, which is a new technology innovation, some researchers argued the causal relationship based on usefulness perception. For instance, He et al. demonstrated that perceived ease of use directly affects perceived usefulness when investigating factors influencing users’ potential adoption intention of autonomous vehicles.⁵⁰ Similarly, Staab confirmed this relationship in a study examining the impact of subjective knowledge on the use of autonomous driving technology.⁵¹ There were some other studies that proposed the point of hedonism perception. They believed that systems that are perceived as easier to use tend to enhance users’ enjoyment.⁵² This relationship has obtained support from Hasan et al., who found that perceived ease of use significantly increases perceived enjoyment in their study on factors shaping online shopping intentions.⁵³ Applying this to AVs, if users perceive the technology as excessively complex, the additional cognitive effort required to operate it may lead to frustration, thereby reducing its hedonic appeal.⁴ Given that perceived risk is an insurmountable and important factor, scholars have specifically demonstrated the impact of technological features on users’ perceived risk and their way of affecting their ultimate adoption intention. Relevant studies have suggested that the ease of interaction with innovative technology also significantly affects consumers’ perception of risk. For instance, Fakhruzzaman and Dimitrova found that perceived ease of use plays a crucial role in reducing perceived risk in the context of e-government services adoption among Indonesian citizens.⁵⁴ A higher perceived ease of use enhances trust in AV,⁵¹ which may mitigate concerns about potential risks related to self-driving.⁵⁵ Based on these findings, we propose the following hypothesis:

Perceived ease of use of AV has a positive effect on the intention of AV adoption (H4a), which is mediated by perceived usefulness (H4b), perceived hedonism (H4c), and perceived risk (H4d).

Compatibility refers to the extent to which an innovation aligns with users’ values, prior experiences, and needs.¹² Prior studies have shown that compatibility plays a crucial role in influencing behavioral intention to adopt various technologies. For instance, Abubakari et al. found that compatibility significantly affected users’ intention to adopt digital technologies in Islamic education institutions,⁵⁶ while Luna et al. reported a similar effect in the context of m-payment.⁵⁷ In the domain of autonomous driving, recent research has begun to confirm that when AV technology is perceived as compatible with users’ values and daily routines, it positively influences their willingness to use it.⁵⁸ These findings suggest that improving perceived compatibility may be a key strategy for accelerating the acceptance and adoption of autonomous vehicles.

Existing research has investigated the impact of perceived compatibility on usage intention from various perspectives, with several studies emphasizing the role of perceived usefulness. For instance, Singh and Sinha found that perceived compatibility significantly influences perceived usefulness in their study of merchants’ adoption of mobile wallet technology.⁵⁹ Similar findings have emerged in autonomous driving research. Yuen et al. argued that autonomous vehicles perceived as compatible with drivers’ needs and values are more likely to be seen as useful.¹⁴ Other studies have approached the relationship from the perspective of hedonic perception, suggesting that compatibility shapes consumers’ emotional responses to innovation.⁶⁰ Recent research on mobile payments further supports this view, indicating that perceived compatibility directly enhances perceived hedonic value.⁶¹ In the context of AVs, when the technology aligns with users’ experiences, needs, and values, it enhances the driving experience and increases perceived hedonism. Rogers also suggested that compatibility reduces uncertainty, which refers to the difficulty in predicting the outcomes of adopting a new technology.⁶² Perceived risk arises from this uncertainty.⁶³ Accordingly, individuals may assess this uncertainty based on personal experiences, preferences, values, and needs, forming subjective expectations of the associated risks. This hypothesis is supported by Xie and An, who found that compatibility negatively influences perceived risk in their study of new energy vehicles.⁶⁴ Similarly, in autonomous driving, we hypothesize that higher compatibility of AV technology with users will reduce concerns regarding safety and technological failure. Based on these insights, we propose the following hypothesis:

Compatibility of AV has a positive effect on the intention of AV adoption (H5a), which is mediated by perceived usefulness (H5b), perceived hedonism (H5c), and perceived risk (H5d).

Relative advantage refers to the extent to which an innovation is perceived as superior to alternatives.¹² Prior research has demonstrated its direct impact on technology adoption. For instance, Mombeuil et al. found that the perceived relative advantage of WeChat Pay significantly influenced users’ adoption intentions in China.⁶⁵ In the domain of autonomous vehicles, recent research similarly indicates that, compared to conventional cars, the advantages associated with autonomous driving positively affect users’ intention to adopt the technology.^39,66

Existing research suggests that perceived value plays a pivotal role in how relative advantage shapes users’ intention to adopt a technology.⁴ Wang et al. distinguish between perceived usefulness and relative advantage, noting that the former is influenced by the latter.⁶⁷ Specifically, the perceived usefulness of an innovation increases when it is seen as a superior alternative.⁶⁸ In the case of AVs, their perceived usefulness is enhanced when they are viewed as a better alternative to conventional vehicles. Beyond this, the relative advantage of an innovation also triggers positive emotional responses. Moons and De Pelsmacker found that recognizing an innovation’s ability to save time, effort, or provide greater efficiency fosters user satisfaction and enjoyment.⁶¹ Autonomous driving technology, by reducing driver involvement and allowing for greater attention freedom, provides a more relaxed and comfortable travel experience.⁸ This aligns with perceived hedonism, indicating that the relative advantages of AVs not only enhance their functional value but also significantly contribute to perceived hedonic value. Additionally, when people recognize that a technology has clear advantages, they tend to reduce their negative assessments of that technology to maintain internal cognitive consistency, thereby aligning their judgments with their perceptions.^69,70 In other words, the stronger the perceived advantage, the more individuals tend to rationalize their positive attitude toward the technology by perceiving lower risks. Therefore, compared to traditional vehicles, the certainty of the advantages of autonomous vehicles may alleviate the concerns of potential adopters, reduce anxiety, and consequently lower perceived risks. Based on these insights, we propose the following hypothesis:

The relative advantage of AV has a positive effect on the intention of AV adoption (H6a), which is mediated by perceived usefulness (H6b), perceived hedonism (H6c), and perceived risk (H6d).

The moderating role of subjective norms

Subjective norms refer to the extent to which individuals are influenced by the groups around them.⁷¹ People often learn and make decisions on the basis of their observations of others’ behaviors,⁷² meaning that influences from significant others shape an individual’s actions. Such behavior mode is also recognized as the Bandwagon effect. It is more likely to be adopted when people face new phenomena. For example, Wan et al. reported in their study on urban green space (UGS) usage intention in Hong Kong that under strong subjective norm influences, individuals were more likely to use UGS, even when they perceived lower levels of usefulness.⁷³ Similarly, Hammad et al. demonstrated that subjective norms moderate the relationship between perceived usefulness and perceived hedonism, positively influencing healthcare providers’ acceptance of digital marketing.⁷⁴

Additionally, uncertainty identity theory posits that when people experience uncertainty, they identify with specific groups through behaviors such as self-categorization and depersonalization, and predict others’ attitudes and behaviors to reduce feelings of uncertainty. This, in turn, moderates the negative impact of perceived risk on the willingness to adopt innovations.⁷⁵ Bhatti and Akram confirmed this finding by showing that subjective norms significantly moderated the effect of privacy risk on online shopping behavior in Pakistan.⁷² On the basis of these findings, we propose the following hypothesis:

Subjective norms positively moderate the effect of perceived usefulness on the intention of AV adoption.

Subjective norms positively moderate the effect of perceived hedonism on the intention of AV adoption.

Subjective norms positively moderate the effect of perceived risk on the intention of AV adoption.

Data collection

Measurement of variables

The constructs in this study were measured using established scales adapted from prior literature to align with the specific AV context and target population, as well as to suit the research scenarios. Each construct was assessed on a five-point Likert scale, with 1 representing “strongly disagree” and 5 “strongly agree.” Before data collection, the scale items were refined based on expert feedback and interviews with five graduate students. Table 1 presents the finalized constructs, measurement items, and their sources.

Table 1.

Constructs and measurement items.

Construct	ID	Measurement item	References
Perceived ease of use (PEU)	PEU1	Learning to use AVs will be easy for me	Yuen et al. (2021)
	PEU2	I will find AVs easy to use
	PEU3	It will be easy for me to become skillful at using AVs
	PEU4	Interacting with an AV would not require a lot of mental effort
Compatibility (COM)	COM1	Using AVs fits well with my lifestyle	Karahanna et al. (2006), Yuen et al. (2020)
	COM2	AVs would be compatible with my mobility needs
	COM3	AVs would be in line with my everyday life
	COM4	AVs would be compatible with my attitude toward technology and my aspirations for the future
Relative advantage (RA)	RA1	AVs would improve my travel efficiency	Qu et al. (2019), Wang (2022)
	RA2	AVs would enhance driving safety
	RA3	AVs would make my arrival at the destination more comfortable
	RA4	AVs would be more advantageous compared to using conventional cars
Perceived usefulness (PU)	PU1	AVs are very useful in my daily life	Davis et al. (1989)
	PU2	I find AVs to be useful when I’m impaired (e.g., drowsy, drunk)
	PU3	Using AVs will decrease my accident risk
	PU4	Using AVs will relieve my stress from driving
Perceived hedonism (PH)	PH1	Using AVs would be fun	Foroughi et al. (2023)
	PH2	Using AVs would be enjoyable
	PH3	Using AVs would be very entertaining
Perceived risk (PR)	PR1	I’m worried that AVs will collect too much personal information from me	Zhang et al. (2019)
	PR2	I’m worried that the failure or malfunctions of AVs may cause accidents
	PR3	I feel that AVs may not perform well, leading to poor usage outcomes
	PR4	I’m worried about the general safety of such technology
Intention of AV adoption (IA)	IA1	I predict I would use AVs in the future	Venkatesh et al. (2012), Zhang et al. (2019)
	IA2	I plan to use AVs in the future
	IA3	If I have enough money, I would purchase AVs in the future
Subjective norms (SN)	SN1	Families/friends/colleagues think it is right to adopt AVs	Si et al. (2024)
	SN2	Somebody important to me recommends adopting AVs
	SN3	If people who influence my behavior think AVs are a good option, I would like to try them
	SN4	If most people around me adopt AVs, I would like to try them

Sample and description

This study employed a questionnaire survey method to collect data from a diverse sample. A total of 748 responses were obtained through three channels: (1) commissioning a professional market research platform for random distribution; (2) utilizing social media for random sampling; and (3) leveraging personal networks via snowball sampling, initiated through acquaintances with driving experience. After excluding invalid responses, 504 valid responses were retained. The questionnaire consisted of three sections. The first section outlined the study’s purpose and objectives and defined the automation level of AVs (i.e., SAE 3 and above) to ensure consistent understanding among participants. It also emphasized the anonymity and confidentiality of responses. The second section gathered demographic data, including age, gender, education level, driving experience, and frequency. The third section addressed the constructs examined in the study, as detailed in Table 1.

Table 2 presents the demographic profile of the sample. Among the respondents, 46.2% were male and 53.8% were female. The majority (86.7%) were under 40 years old, a demographic group with significant potential as future consumers of AVs, providing valuable insights for market forecasting. Regarding education, 90.1% held a bachelor’s degree or higher, which is associated with greater capacity for understanding and providing detailed feedback. Additionally, 57.7% had over 3 years of driving experience, and 66.4% drove more than four times per week, indicating extensive experience with traditional driving.

Table 2.

Respondents’ profile.

Characteristics	Items	Frequency (n = 504)	Percentage (%)
Gender	Male	233	46.2
Gender	Female	271	53.8
Age (years)	18–29	149	29.6
	30–39	288	57.1
	40–49	48	9.5
	50–60	19	3.8
Education level	High school or lower	13	2.6
	Diploma	37	7.3
	Bachelor	406	80.6
	Postgraduate or above	48	9.5
Driving experience (years)	<1	21	4.2
	1–3	96	19
	4–6	179	35.5
	7–9	112	22.2
	≥10	96	19
Driving frequency	<1 times/month	21	4.2
	2–4 times/month	148	29.4
	5–7 times/month	231	45.8
	>7 times/month	104	20.6

Data analysis and results

This study employed a regression method to analyze the data.⁷⁶ The analysis proceeded in three stages: first, an exploratory factor analysis identified the structure of the latent variables; second, a confirmatory factor analysis assessed the model’s reliability and validity, alongside a test for common method bias; and third, a multiple regression analysis was performed through path analysis to examine the relationships among variables. Data analysis was conducted using SPSS version 23.

Measurement model testing

Principal component analysis (PCA) was first conducted on the constructs. Before PCA, the Kaiser–Meyer–Olkin (KMO) measure was used to assess the suitability of the data for factor analysis. The KMO value of 0.836 and the p-value of Bartlett’s test of sphericity (p < 0.001) indicated that the data were appropriate for factor analysis. The factor analysis extracted eight factors, accounting for 67.174% of the total variance. Rotated factor loadings for each item are presented in Table 3. Except for item SN4, all loadings exceeded 0.7, suggesting satisfactory convergent validity. Consequently, SN4 was removed from the measurement model.⁷⁷

Table 3.

Exploratory factor analysis.

	PEU	COM	RA	PU	PH	PR	IA	SN
PEU1	0.817	0.049	−0.015	0.102	0.139	−0.052	0.075	−0.025
PEU2	0.831	0.146	0.012	0.046	0.059	−0.05	0.138	−0.018
PEU3	0.824	0.128	0.032	0.069	0.071	−0.016	0.144	−0.025
PEU4	0.802	0.069	0.009	0.026	0.023	−0.02	0.18	0.042
COM1	0.101	0.811	0.029	0.108	0.065	−0.067	0.076	−0.019
COM2	0.096	0.818	0.043	0.065	0.081	−0.135	0.037	0.053
COM3	0.069	0.721	−0.081	0.125	0.139	−0.181	0.128	0
COM4	0.124	0.737	0.008	0.111	0.085	−0.143	0.129	0.015
RA1	−0.025	0.023	0.777	0.118	0.016	−0.089	−0.122	−0.043
RA2	0.036	0.028	0.797	0.061	0.001	−0.048	0.065	0.034
RA3	0.006	−0.043	0.809	0.107	0.104	−0.059	0.061	−0.015
RA4	0.017	−0.001	0.781	0.028	0.063	−0.052	0.126	−0.03
PU1	0.094	0.064	0.087	0.801	0.079	−0.001	0.108	−0.016
PU2	0.048	0.115	0.09	0.763	−0.082	−0.106	0.005	−0.002
PU3	0.043	0.113	0.038	0.770	0.047	0.004	0.093	−0.055
PU4	0.044	0.08	0.101	0.776	−0.019	−0.028	0.071	0.047
PH1	0.066	0.179	0.040	0.024	0.782	−0.043	0.024	0.115
PH2	0.087	0.102	0.109	−0.036	0.765	−0.039	0.078	0.159
PH3	0.115	0.059	0.037	0.033	0.823	−0.076	0.047	0.063
PR1	−0.041	−0.155	−0.011	−0.042	−0.048	0.848	−0.043	−0.037
PR2	−0.046	−0.108	−0.105	0.014	−0.014	0.805	−0.125	−0.117
PR3	−0.031	−0.134	−0.095	−0.042	−0.079	0.799	−0.142	−0.136
PR4	−0.023	−0.12	−0.066	−0.069	−0.043	0.816	−0.136	−0.016
IA1	0.205	0.152	0.012	0.089	0.079	−0.158	0.760	0.078
IA2	0.182	0.171	0.08	0.155	0.04	−0.108	0.788	0.102
IA3	0.224	0.073	0.072	0.086	0.057	−0.221	0.811	0.085
SN1	0.025	0.056	−0.023	0.058	0.02	−0.048	−0.039	0.808
SN2	0.064	0.02	0.009	−0.01	0.023	−0.09	−0.053	0.834
SN3	−0.076	0.006	−0.015	0.005	0.125	−0.027	0.188	0.711
SN4	−0.039	−0.042	−0.025	−0.087	0.196	−0.113	0.164	0.583

Bold values represent the highest factor loadings of each item on the factor it is most strongly associated with.

Subsequently, a validation factor analysis was performed on the constructs after the removal of specific items to assess the reliability and validity of the measurement model. As shown in Table 4, the Cronbach’s α and composite reliability (CR) values for all constructs exceeded 0.7, indicating strong reliability.⁷⁷ Most standardized factor loadings met the desired threshold of 0.7; although some values fell within the 0.5–0.7 range, they remained above the acceptable minimum of 0.5.⁷⁸ Furthermore, the average variance extracted (AVE) for all constructs exceeded 0.5, signifying that more than 50% of the variance was attributable to the constructs themselves rather than measurement error, thus confirming the convergent validity of the model.⁷⁷

Table 4.

Confirmatory factor analysis results.

Construct	Item	$λ$	AVE	CR	Cronbach’s α
Perceived ease of use (PEU)	PEU1	0.764	0.615	0.865	0.861
	PEU2	0.816
	PEU3	0.805
	PEU4	0.750
Compatibility (COM)	COM1	0.755	0.540	0.824	0.820
	COM2	0.776
	COM3	0.700
	COM4	0.706
Relative advantage (RA)	RA1	0.690	0.523	0.814	0.812
	RA2	0.717
	RA3	0.781
	RA4	0.700
Perceived usefulness (PU)	PU1	0.762	0.509	0.806	0.804
	PU2	0.683
	PU3	0.700
	PU4	0.707
Perceived hedonism (PH)	PE1	0.718	0.513	0.759	0.758
	PE2	0.707
	PE3	0.723
Perceived risk (PR)	PR1	0.798	0.618	0.866	0.865
	PR2	0.769
	PR3	0.797
	PR4	0.781
Intention of AV adoption (IA)	IA1	0.729	0.625	0.833	0.829
	IA2	0.786
	IA3	0.852
Subjective norms (SN)	SN1	0.745	0.509	0.751	0.738
	SN2	0.825
	SN3	0.539

Finally, Table 5 shows the correlation coefficients between the constructs and the square root of the AVE. The square root of the AVE for all the constructs is greater than the value of the correlation coefficient between that construct and any other construct, a result that indicates good discriminant validity between the constructs in this study.

Table 5.

Discriminant validity.

	PEU	COM	RA	PU	PH	PR	IA	SN
PEU	0.784
COM	0.273**	0.735
RA	0.032	0.037	0.723
PU	0.173**	0.264**	0.206**	0.714
PH	0.211**	0.269**	0.126**	0.055	0.716
PR	−0.128**	−0.336**	−0.160**	−0.119**	−0.167**	0.786
IA	0.409**	0.329**	0.109*	0.249**	0.209**	−0.358**	0.790
SN	0.019	0.067	−0.010	0.012	0.219**	−0.171**	0.167**	0.710

^aThe square roots of AVEs are along the main diagonal.

^bCorrelations between constructs are below the main diagonal.

Common method bias

To mitigate potential common method bias, several procedures were implemented. First, the questionnaire was designed to separate dependent and independent measures, minimizing sequencing effects. Second, Harman’s single-factor analysis (HSA) was conducted to determine whether a single factor could account for the total variance in the constructs. The results showed that the unrotated exploratory factor analysis extracted eight factors with eigenvalues greater than 1, and the maximum variance explained by a single factor was 20.48%, well below the 40% threshold, indicating no substantial common method bias. Additionally, a common latent factor was incorporated into the original measurement model (Model M2), which was compared with the original model (Model M1). Model fit indices, presented in Table 6, demonstrated that the differences in CFI, TLI, and RMSEA between Models M1 and M2 were all below the 0.05 threshold,⁷⁹ further confirming the absence of significant common method bias.

Table 6.

Results of common method bias.

	$χ 2$	df	$CMIN / DF$	CFI	TLI	RMSEA
M1	471.173	377	1.250	0.984	0.982	0.022
M2	404.84	348	1.163	0.99	0.988	0.018

Regression analysis

Table 7 shows the fit of the structural model, from which it can be seen that all the goodness-of-fit metrics are within the recommended range of reference values, indicating that the model is well-fitted.⁷⁷

Table 7.

Structural model assessment.

	$CMIN / DF$	CFI	TLI	RMSEA
Results	1.695	0.945	0.939	0.037
Standards	Between 1 and 3	>0.9	>0.9	<0.08

Table 8 shows the regression results between determinants and the intention of AV adoption. Model 1 illustrates the causal relationships between customer value perceptions and the intention of AV adoption, while Model 2 illustrates the relationships between technology innovation characteristics and the intention of AV adoption. Before observing the coefficients, we checked the multicollinearity of factors illustrated by the variance inflation factor (VIF). It reveals no such threats in each regression model. Results of Model 1 demonstrate that perceived usefulness, perceived hedonism, and perceived risk significantly influence users’ willingness to use, with standardized effects of 0.204, 0.150, and −0.309, respectively. Thus, H1, H2, and H3 were all supported by empirical data. Moreover, by comparing the coefficient values of the three variables, we can see, perceived risk plays the most important determinant role on the customer’s intention of AV adoption. Results of Model 2 reveal that perceived ease of use, compatibility, and relative advantage all have significant positive effects on AV adoption intention, with standardized effects of 0.339, 0.229, and 0.104, respectively. Thus, H4a, H5a, and H6a were all supported.

Table 8.

Regression analysis results.

Variables	IA		VIF
Variables	Model 1	Model 2	VIF
Independent variable
PU	0.204^***		1.020
PH	0.150^***		1.046
PR	−0.309^***		1.070
PEU		0.339^***	1.088
COM		0.229^***	1.092
RA		0.104^**	1.018
Control variable
Gender	0.011	−0.010
Age	−0.003	−0.002
Education	0.018	0.001
Experience of driving	−0.052	0.002
Frequency of driving	0.084	0.051

Note: PU: perceived usefulness; IA: intention of AV adoption; PH: perceived hedonism; PR: perceived risk; PEU: perceived ease of use; COM: compatibility; RA: relative advantage.

Table 9 shows the mediating test results using software of PROCESS. The findings reveal that perceived ease of use (PEU), compatibility (COM), and relative advantage (RA) significantly influence users’ intention to adopt autonomous vehicles (IA) via perceived usefulness (PU), perceived hedonic value (PH), and perceived risk (PR).

Table 9.

Results of the mediating effect.

	Pathway	Effect	SE	LLCI	ULCI
PEU	PEU → PU → IA	0.026	0.010	0.006	0.047
	PEU → PH → IA	0.018	0.008	0.002	0.032
	PEU → PR → IA	0.032	0.014	0.007	0.06
	Direct effect: PEU → IA	0.314	0.040	0.246	0.394
	Total effect: PEU → IA	0.390	0.040	0.315	0.471
COM	COM → PU → IA	0.049	0.015	0.026	0.079
	COM → PH → IA	0.033	0.013	0.011	0.062
	COM → PR → IA	0.094	0.019	0.060	0.129
	Direct effect: COM → IA	0.179	0.047	0.071	0.253
	Total effect: COM → IA	0.356	0.047	0.264	0.448
RA	RA → PU → IA	0.043	0.013	0.021	0.067
	RA → PH → IA	0.021	0.010	0.005	0.038
	RA → PR → IA	0.049	0.016	0.020	0.085
	Direct effect: RA → IA	0.014	0.041	−0.066	0.078
	Total effect: RA → IA	0.126	0.047	0.035	0.218

Specifically, all three indirect paths from PEU to IA are significant, with none of the confidence intervals including zero. Among these, the path through PR exhibits the strongest mediating effect (β = 0.032), followed by PU (β = 0.026) and PH (β = 0.018). The direct effect of PEU on IA remains significant (β = 0.314), indicating partial mediation. Similarly, COM significantly affects IA through PU (β = 0.049), PH (β = 0.033), and PR (β = 0.094), with all indirect effects being statistically significant. The direct effect of COM also remains significant (β = 0.179), further supporting partial mediation. In contrast, RA exerts a significant influence on IA exclusively through the mediators—PU, PH, and PR—while its direct effect is not significant (CI: −0.066 to 0.078). This highlights the critical mediating roles played by these three perceptual factors in the relationship between RA and IA.

Table 10 presents the results of the moderation test for subjective norms. The findings indicate that subjective norms significantly moderate only the relationship between perceived risk and willingness to use, thus supporting H9 while rejecting H7 and H8. To illustrate the moderating effect more clearly, Figure 2 depicts the interaction between perceived risk and willingness to use. As shown, higher levels of subjective norms reduce the slope of the regression line, suggesting that stronger subjective norms mitigate the negative impact of perceived risk on willingness to use.

Table 10.

Results of the moderating effect.

Hypothesis	Path	$β$	SE	T	P	95% CI
H7	PU*SN→ IA	−0.0477	0.0587	−0.8121	0.4171	(−0.1631, 0.0677)
H8	PH*SN→ IA	−0.0072	0.0555	−0.1290	0.8974	(−0.1163, 0.1020)
H9	PR*SN→ IA	0.1107	0.0527	2.0991	0.0363	(0.0071, 0.2143)

Figure 2.

The interaction effect of perceived risk and subjective norms on AV adoption intention.

Conclusions

Main findings and discussions

First, the results validate the significant effects of perceived usefulness (PU), perceived hedonism (PH), and perceived risk (PR) on users’ willingness to adopt autonomous vehicles (AVs). Specifically, PU (β = 0.204, p < 0.001) and PH (β = 0.150, p < 0.001) positively influence adoption willingness, indicating that users are more likely to adopt AVs if they perceive them as useful or enjoyable. In contrast, PR negatively impacts adoption (β = −0.309, p < 0.001), suggesting that concerns about safety, control, and technical reliability significantly hinder adoption intentions. Notably, PR exerts a stronger negative effect on adoption willingness than the positive influences of PU and PH. This implies that, in the case of high-risk technologies like AVs, consumers’ concerns about potential risks may serve as a major barrier to widespread adoption, even if the technology is seen as beneficial or enjoyable. This contrasts with Tan et al., who found that higher PR was associated with greater general acceptance of fully automated vehicles (FAVs).⁸⁰ In their study, perceived benefits were stronger predictors of acceptance than PRs. This discrepancy may stem from the fact that the current study examines Level 3 AVs, which still require some driver involvement, whereas Tan et al. focused on fully automated vehicles that require no driver intervention. Consequently, public perceptions of risk and acceptance may vary across different levels of automation. Future research should investigate the mechanisms of PR at varying automation levels and its specific impact on adoption willingness.

Second, the findings confirm that key innovation characteristics—PEU, COM, and RA—positively influence users’ intention to adopt AVs, aligning with prior research on innovation diffusion and technology acceptance. Further analysis reveals that these characteristics influence behavioral intention primarily through indirect pathways mediated by PU, PH, and PR, rather than through direct effects alone. This underscores the critical role of users’ value perceptions in shaping adoption behavior, particularly in the case of RA, whose influence on adoption intention operates largely through these mediators. Among the mediators, PR consistently emerges as the strongest negative pathway across all models, further highlighting that safety concerns and perceived uncertainty are key barriers to adoption. Even when the AV technology demonstrates strong functional advantages, psychological perceptions of risk may still attenuate its positive effects on user intention.

Finally, subjective norms significantly moderate the relationship between PR and IA (β = 0.1107, p = 0.0363), but not the relationships between PH or PU and adoption willingness. Specifically, when subjective norms support innovation adoption, they weaken the negative effect of PR on adoption and increase adoption intentions. Conversely, when subjective norms oppose adoption, they amplify the negative effect of PR, inhibiting adoption intentions. This finding underscores that, in the context of emerging technologies such as autonomous driving, users’ perceptions of risk are not only shaped by personal concerns about technological failure, safety, and other factors but are also influenced by social attitudes toward the technology. In contrast, PU and PH are more closely linked to individual evaluations of the technology’s value and experience, often driven by personal needs and direct experiences rather than external social influences. As a result, the impact of subjective norms is weaker on PU and PH.

Theoretical and practical value

This study makes several theoretical contributions, advancing our understanding of AV adoption. Firstly, it systematically constructs a value assessment framework to capture how consumers evaluate autonomous driving technology. By integrating user motivations—perceived usefulness and perceived hedonism—with perceived risks, the study constructs a dual psychological mechanism comprising perceived benefits and perceived risks. And it is shown that concerns about risks are most dominant when consumers make adoption decisions of AVs, compared with potential benefits such as functional usefulness and hedonic value. This integrated framework provides a more holistic understanding of the cognitive processes that shape consumer attitudes toward AVs, thereby enriching the theoretical foundation of technology acceptance research.

Secondly, by analyzing the direct impact of perceived innovation characteristics on adoption intention, this study identifies the core technological drivers of autonomous vehicle diffusion, thereby extending the application of IDT within the autonomous driving context. Among the three focused features, ease of use has the most dominant impact on adoption intention. It demonstrates that the technology innovation of AVs should pay more attention to the ease of use function, compared with compatibility and relative advantages.

Furthermore, the study provides empirical evidence that perceived value serves as a critical mediating variable in the relationship between innovation characteristics and adoption intention. This finding sheds light on the cognitive mechanisms through which consumers interpret the innovation value of AVs, and it delineates a complete psychological pathway from the initial perception of technological attributes to the eventual formation of behavioral intentions. In doing so, it offers a novel theoretical perspective on how consumers accept and adopt emerging technologies in the marketplace.

From a practical perspective, the findings of this study provide critical insights for relevant stakeholders. Perceived risk has been identified as a major barrier to the adoption of autonomous driving technologies. Accordingly, regulators should prioritize the development of robust safety standards and the enforcement of industry regulations to strengthen public trust in such technologies. Moreover, the significant influence of compatibility and relative advantage on user adoption underscores the need for manufacturers to align product design and marketing strategies with users’ values and lifestyle preferences. Furthermore, positive social norms have been shown to mitigate risk perception, highlighting the importance of sociocultural and public opinion factors in promoting the diffusion of autonomous vehicles. Effective dissemination strategies should thus incorporate community engagement and leverage the influence of social role models and opinion leaders to foster a supportive social climate and enhance adoption willingness.

Footnotes

Acknowledgements

All authors who have made substantial contributions to this work are reported in the manuscript. Each author certifies that they have participated sufficiently in the work and take responsibility for its content, including participation in the concept, design, analysis, writing, and revision of the manuscript. Furthermore, each author certifies that this material has not been and will not be submitted to or published in any other publication before its appearance in Human Systems Management.

ORCID iDs

Wei Li

Yuyue Liu

Ethical considerations

This study did not require formal ethical approval, as it involved an anonymous survey with voluntary participation.

Consent for publication

Informed consent was obtained from all respondents before data collection.

Author contributions

CONCEPTION: Wei Li, Yuyue Liu; METHODOLOGY: Yuyue Liu; DATA COLLECTION: Wei Li, Yuyue Liu; INTERPRETATION OR ANALYSIS OF DATA: Yuyue Liu; PREPARATION OF THE MANUSCRIPT: Yuyue Liu; REVISION FOR IMPORTANT INTELLECTUAL CONTENT: Wei Li; SUPERVISION: Wei Li.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Data will be made available on request.*

References

Meyer-Waarden

Cloarec

. “Baby, you can drive my car”: psychological antecedents that drive consumers’ adoption of AI-powered autonomous vehicles. Technovation 2022; 109: 102348.

Parekh

Poddar

Rajpurkar

, et al. A review on autonomous vehicles: progress, methods and challenges. Electronics 2022; 11: 2162.

Lim

Lee

Sung

, et al. Is Australia ready for autonomous vehicles? examining the factors influencing AV adoption through expert interviews. Australas Market J 2023; 32: 308–322.

Yuen

Wong

, et al. The determinants of public acceptance of autonomous vehicles: an innovation diffusion perspective. J Clean Prod 2020; 270: 121904.

Olayode

Severino

, et al. Systematic literature review on the applications, impacts, and public perceptions of autonomous vehicles in road transportation system. J Traffic Transport Eng 2023; 10: 1037–1060.

Garikapati

Shetiya

. Autonomous vehicles: evolution of artificial intelligence and the current industry landscape. Big Data Cogn Comput 2024; 8: 42.

Reed

Campbell

Thomas

. The value of autonomous vehicles for last-mile deliveries in urban environments. Manag Sci 2022; 68: 280–299.

Chen

Yang

Luan

, et al. Advanced autonomous vehicle acceptance model (AAVAM) for promoting accessible travel. In: 2024 29th international conference on automation and computing (ICAC), 2024, pp. 1–6: IEEE.

Gaio

Cugurullo

. Cyclists and autonomous vehicles at odds: can the transport oppression cycle be broken in the era of artificial intelligence? AI Soc 2023; 38: 1223–1237.

10.

Zefreh

Edries

Esztergár-Kiss

, et al.

Intention to use private autonomous vehicles in developed and developing countries: what are the differences among the influential factors, mediators, and moderators?

Travel Behav Soc 2023; 32: 100592.

11.

Naiseh

Clark

Akarsu

, et al. Trust, risk perception, and intention to use autonomous vehicles: an interdisciplinary bibliometric review. AI Soc 2024; 40: 1091–1111.

12.

Rogers

. Diffusion of innovations: modifications of a model for telecommunications. In: Stoetzer

M-W

Mahler

(eds). Die diffusion von innovationen in der telekommunikation. Springer, 1995, pp. 25–38.

13.

Moore

Benbasat

. Development of an instrument to measure the perceptions of adopting an information technology innovation. Inf Syst Res 1991; 2: 192–222.

14.

Yuen

Cai

, et al. Factors influencing autonomous vehicle adoption: an application of the technology acceptance model and innovation diffusion theory. Technol Anal Strat Manag 2021; 33: 505–519.

15.

Yoon

S-J

. Validation of haptic enabling technology acceptance model (HE-TAM): integration of IDT and TAM. Telematics Inf 2014; 31: 585–596.

16.

Mbatha

. Diffusion of innovations: how adoption of new technology spreads in society. In: Ocholla

Onyancha

Adesina

(eds). Information, knowledge, and technology for teaching and research in Africa: human machine interaction and user interfaces. Springer Nature Switzerland, 2024, pp. 1–18.

17.

Takahashi

Figueiredo

JCBD

Scornavacca

. Investigating the diffusion of innovation: a comprehensive study of successive diffusion processes through analysis of search trends, patent records, and academic publications. Technol Forecast Soc Change 2024; 198: 122991.

18.

Ayanwale

Ndlovu

. Investigating factors of students' behavioral intentions to adopt chatbot technologies in higher education: perspective from expanded diffusion theory of innovation. Comp Human Behav Rep 2024; 14: 100396.

19.

Farzin

Mamdoohi

Ciari

. Autonomous vehicles acceptance: a perceived risk extension of unified theory of acceptance and use of technology and diffusion of innovation, evidence from Tehran, Iran. Int J Hum Comput Interact 2022; 39: 2663–2672.

20.

Hasan Emon

. Insights into technology adoption: a systematic review of framework, variables and items. Inform Manag Comp Sci 2023; 6: 55–61.

21.

Ryan

Deci

. Intrinsic and extrinsic motivations: classic definitions and new directions. Contemp Educ Psychol 2000; 25: 54–67.

22.

Davis

Granić

. Evolution of TAM. In: The technology acceptance model. Springer International Publishing, 2024, pp. 19–57.

23.

Venkatesh

Thong

JYL

. Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology. MIS Q 2012; 36: 157–178.

24.

Seuwou

. Measuring behavioural intention to accept autonomous vehicles: a structural equation model. PhD. London South Bank University, 2020.

25.

Hegner

Beldad

Brunswick

. In automatic we trust: investigating the impact of trust, control, personality characteristics, and extrinsic and intrinsic motivations on the acceptance of autonomous vehicles. Int J Hum Comput Interact 2019; 35: 1769–1780.

26.

Kieu

Senanayake

. Perception, experience and resilience to risks: a global analysis. Sci Rep 2023; 13: 19356.

27.

Habbal

Ali

Abuzaraida

. Artificial intelligence trust, risk and security management (AI TRiSM): frameworks, applications, challenges and future research directions. Expert Syst Appl 2024; 240: 122442.

28.

Benleulmi

Ramdani

. Behavioural intention to use fully autonomous vehicles: instrumental, symbolic, and affective motives. Transport Res F Traffic Psychol Behav 2022; 86: 226–237.

29.

Man

Xiong

Chang

, et al. Critical factors influencing acceptance of automated vehicles by Hong Kong drivers. IEEE Access 2020; 8: 109845–109856.

30.

Alam

. Data privacy and security in autonomous connected vehicles in smart city environment. Big Data Cogn Comput 2024; 8: 95.

31.

Lin

K-Y

H-P

. Why people use social networking sites: an empirical study integrating network externalities and motivation theory. Comput Hum Behav 2011; 27: 1152–1161.

32.

C-G

Lee

C-S

, et al. Factors affecting the behavioral intention to adopt mobile banking: an international comparison. Technol Soc 2020; 63: 101360.

33.

Roh

Seok

Kim

. Unveiling ways to reach organic purchase: green perceived value, perceived knowledge, attitude, subjective norm, and trust. J Retailing Consum Serv 2022; 67: 102988.

34.

Watanabe

Wakui

. Acceptance of main power generation sources among Japan’s undergraduate students: the roles of knowledge, experience, trust, and perceived risk and benefit. Sustainability 2021; 13: 12416.

35.

Featherman

Jia

Califf

, et al. The impact of new technologies on consumers beliefs: reducing the perceived risks of electric vehicle adoption. Technol Forecast Soc Change 2021; 169: 120847.

36.

Sohn

Kwon

. Technology acceptance theories and factors influencing artificial intelligence-based intelligent products. Telematics Inf 2020; 47: 101324.

37.

Tamilmani

Rana

Wamba

, et al. The extended unified theory of acceptance and use of technology (UTAUT2): a systematic literature review and theory evaluation. Int J Inf Manag 2021; 57: 102269.

38.

Park

Hong

. Adopting autonomous vehicles: the moderating effects of demographic variables. J Retailing Consum Serv 2021; 63: 102687.

39.

Xiao

Goulias

. Perceived usefulness and intentions to adopt autonomous vehicles. Transport Res Pol Pract 2022; 161: 170–185.

40.

Keszey

. Behavioural intention to use autonomous vehicles: systematic review and empirical extension. Transport Res C Emerg Technol 2020; 119: 102732.

41.

Huang

. Psychological factors affecting potential users’ intention to use autonomous vehicles. PLoS One 2023; 18: e0282915.

42.

Foroughi

Nhan

Iranmanesh

, et al. Determinants of intention to use autonomous vehicles: findings from PLS-SEM and ANFIS. J Retailing Consum Serv 2023; 70: 103158.

43.

Mayer

Davis

Schoorman

. An integrative model of organizational trust. Acad Manag Rev 1995; 20: 709–734.

44.

Kapser

Abdelrahman

. Acceptance of autonomous delivery vehicles for last-mile delivery in Germany – extending UTAUT2 with risk perceptions. Transport Res C Emerg Technol 2020; 111: 210–225.

45.

Zhang

Tao

, et al. The roles of initial trust and perceived risk in public’s acceptance of automated vehicles. Transp Res Part C 2019; 98: 207–220.

46.

Waung

McAuslan

Lakshmanan

. Trust and intention to use autonomous vehicles: manufacturer focus and passenger control. Transport Res F Traffic Psychol Behav 2021; 80: 328–340.

47.

Venkatesh

Davis

. A model of the antecedents of perceived ease of use: development and test. Decis Sci J 1996; 27: 451–481.

48.

Wang

, et al. Exploring user adoption of ChatGPT: a technology acceptance model perspective. Int J Hum Comput Interact 2024; 41: 1431–1445.

49.

Pillai

Sivathanu

Metri

, et al. Students’ adoption of AI-based teacher-bots (T-bots) for learning in higher education. Inf Technol People 2023; 37: 328–355.

50.

Zhang

Guo

, et al. Understanding the adoption of autonomous vehicles in China based on TRI and TAM. World Electric Vehicle Journal 2025; 16: 23.

51.

Staab

Liebherr

. Perception and acceptance of autonomous vehicles: influencing factors and the relevance of subjective knowledge. Int J Hum Comp Inter 2024; 40: 8463–8475.

52.

C-H

Huey-Min

Yuan-Chi

, et al. Effects of user and system characteristics on perceived usefulness and perceived ease of use of the web-based classroom response system. Turkish Online J Educ Techn-TOJET 2012; 11(3): 128–143.

53.

Hasan

AA-T

Sumon

Islam

, et al. Factors influencing online shopping intentions: the mediating role of perceived enjoyment. Tujom 2021; 6: 239–253.

54.

Fakhruzzaman

Dimitrova

. Factors influencing e-government adoption in Indonesia: the importance of perceived risk. Control Syst 2020; 12: 125–131.

55.

Kenesei

Ásványi

Kökény

, et al. Trust and perceived risk: how different manifestations affect the adoption of autonomous vehicles. Transport Res Pol Pract 2022; 164: 379–393.

56.

Abubakari

Zakaria

Musa

. Perceived compatibility and students’ intention to adopt digital technologies in Islamic education institutions. Cogent Educ 2024; 11: 2430869.

57.

Ramos de Luna

Montoro-Ríos

Molinillo

, et al. Consumer behaviour and mobile payments in the point of Sale: exploring the determinants of intention to adopt it. Int J Hum Comput Interact 2023; 40: 5350–5372.

58.

Mansoori

Al-Emran

Shaalan

. Factors affecting autonomous vehicles adoption: a systematic review, proposed framework, and future roadmap. Int J Hum Comp Inter 2024; 40: 8397–8418.

59.

Singh

Sinha

. How perceived trust mediates merchant’s intention to use a mobile wallet technology. J Retailing Consum Serv 2020; 52: 101894.

60.

Moons

De Pelsmacker

. An extended decomposed theory of planned behaviour to predict the usage intention of the electric car: a multi-group comparison. Sustainability 2015; 7: 6212–6245.

61.

Wang

ES-T

. Influences of innovation attributes on value perceptions and usage intentions of mobile payment. J Elec Comm Res 2022; 23: 45–48.

62.

Childers

Rao

. The influence of familial and peer-based reference groups on consumer decisions. J Consum Res 1992; 19: 198–211.

63.

Vilko

Tariq

Nemilentseva

, et al. Managing uncertainty in innovation process through AI integration. In: Rocha

Montenegro

Pereira

, et al. (eds). Management, tourism and smart technologies. Springer Nature Switzerland, 2024, pp. 153–161.

64.

Xie

. Motivation to interaction media: the impact of automation trust and self-determination theory on intention to use the new interaction technology in autonomous vehicles. Morden Economic Sci 2020; 42(05): 113–121.

65.

Mombeuil

Uhde

. Relative convenience, relative advantage, perceived security, perceived privacy, and continuous use intention of China€s WeChat pay: a mixed-method two-phase design study. J Retailing Consum Serv 2021; 59: 102384.

66.

Yuen

Chua

Wang

, et al. Understanding public acceptance of autonomous vehicles using the theory of planned behaviour. Int J Environ Res Publ Health 2020; 17: 4419.

67.

Wang

Meister

Wang

. Relative advantage and perceived usefulness: the adoption of competing ICTs. DIGIT 2008 Proc 2008; 6: 1–21.

68.

Yuan

, et al. Proposing a multifaceted model for adopting prefabricated construction technology in the construction industry. Eng Construct Architect Manag 2021; 30: 755–786.

69.

Marikyan

Papagiannidis

Alamanos

. Cognitive dissonance in technology adoption: a study of smart home users. Inf Syst Front 2023; 25: 1101–1123.

70.

Hsieh

C-C

Kuo

P-L

Yang

S-C

, et al. Assessing blog-user satisfaction using the expectation and disconfirmation approach. Comput Hum Behav 2010; 26: 1434–1444.

71.

Mustofa

Kuncoro

Atmono

, et al. Extending the technology acceptance model: the role of subjective norms, ethics, and trust in AI tool adoption among students. Comput Educ Artif Intell 2025; 8: 100379.

72.

Bhatti

Akram

. The moderating role of subjective norms between online shopping behaviour and its determinants. Int J Soc Sci Eco Review 2020; 2: 1–9.

73.

Wan

Shen

Choi

. The moderating effect of subjective norm in predicting intention to use urban green spaces: a study of Hong Kong. Sustain Cities Soc 2018; 37: 288–297.

74.

Hammad

Bataineh

Alshurideh

, et al. Factors affecting healthcare providers to accept digital marketing: the moderating role of subjective norms. Int J Data Net 2022; 6: 1085–1098.

75.

Grant

Hogg

. Self-uncertainty, social identity prominence and group identification. J Exp Soc Psychol 2012; 48: 538–542.

76.

Hair

Gabriel

Patel

. AMOS covariance-based structural equation modeling (CB-SEM): guidelines on its application as a marketing research tool. REMark, Rev Bras Mark 2014; 13: 44–55.

77.

Hair

Risher

Sarstedt

, et al. When to use and how to report the results of PLS-SEM. Eur Bus Rev 2019; 31: 2–24.

78.

Cai

Yuen

Wang

. Explore public acceptance of autonomous buses: an integrated model of UTAUT, TTF and trust. Travel Behaviour and Society 2023; 31: 120–130.

79.

Califf

Sarker

. The bright and dark sides of technostress: a mixed-methods study involving healthcare IT. MIS Q 2020; 44: 809–856.

80.

Tan

Zhao

Yang

. Exploring the influence of anxiety, pleasure and subjective knowledge on public acceptance of fully autonomous vehicles. Comput Hum Behav 2022; 131: 107187.